Executive Summary
Key findings from the Titanic survival factor analysis
Of 600 passengers analyzed, 226 survived (37.7% overall survival rate). The gender gap was dramatic: 69.6% of female passengers survived versus only 21.2% of males, confirming the 'women and children first' protocol was followed. A logistic regression model achieved an AUC of 0.805, indicating strong predictive discrimination across passenger class, sex, age, and fare.
Survival Rate by Passenger Class
Raw survival rate and survivor count for each ticket class
Survival rates differed sharply by passenger class: Class 1: 57.8%, Class 2: 43.1%, Class 3: 25.9%. First-class passengers had the highest survival probability, reflecting both greater access to lifeboats on upper decks and the social hierarchy of the evacuation. Third-class passengers faced the steepest disadvantage, with a survival rate roughly 32 percentage points below first class.
Survival Rate by Sex
Survival rate and survivor count broken down by sex
69.6% of female passengers survived compared to only 21.2% of male passengers — a gap of 48.4 percentage points. Female passengers (142 survivors) were far more likely to be placed in lifeboats under the 'women and children first' evacuation order. This gender difference is the single strongest predictor in the logistic regression model.
Survival Rate by Class and Sex
Heatmap of survival rates across all passenger class and sex combinations
The heatmap reveals that Class 1 female passengers had the highest survival rate (96.1%) while Class 3 male passengers fared the worst (13.7%). The class gradient is visible in both sexes, but female passengers consistently outperformed their male counterparts within every class. The largest survival gap by sex appears in first class, where evacuation priority most strongly protected women.
Fare Distribution by Survival Status
Box plot comparing fare distributions for survivors and non-survivors
Passengers who survived paid a median fare of £23.06, compared to £16.20 for those who did not — a 1.4x difference. This reflects the strong correlation between fare, cabin deck, and proximity to lifeboats. The distribution among survivors is right-skewed, driven by a small number of very high-fare first-class passengers who almost universally survived.
Logistic Regression: Odds Ratios by Factor
Odds ratios from logistic regression — values below 1 indicate decreased survival odds
Controlling for all other factors simultaneously, sex is the strongest predictor: being male is associated with an odds ratio of 0.10 relative to female, meaning male passengers had dramatically lower survival odds. Higher passenger class numbers (Class 2, Class 3) are also associated with odds ratios below 1 compared to first class, confirming the socioeconomic gradient. Age and fare add incremental explanatory power once class and sex are controlled.
Survival Profile by Demographic Group
Per-group summary of fare, survivors, and survival rate across class-sex combinations
| group | avg_fare | survivors | survival_rate | total_passengers |
|---|---|---|---|---|
| Class 1 / Female | 84.57 | 49 | 0.9608 | 51 |
| Class 2 / Female | 20.49 | 41 | 0.7321 | 56 |
| Class 3 / Female | 14.22 | 52 | 0.5361 | 97 |
| Class 1 / Male | 85.93 | 36 | 0.375 | 96 |
| Class 2 / Male | 19.96 | 18 | 0.2222 | 81 |
| Class 3 / Male | 12.83 | 30 | 0.137 | 219 |
Across all class-and-sex combinations, Class 1 / Female had the highest survival rate (96.1%) while Class 3 / Male had the lowest (13.7%). Average fare tracks closely with survival advantage: higher-fare groups consistently show better outcomes, reflecting the compounding benefits of first-class travel. This table enables direct comparison of survivor counts, group sizes, and economic indicators side by side.
Overall Survival Statistics
Headline statistics from the analysis including model performance metrics
| value | metric |
|---|---|
| 600 | Total Passengers |
| 226 | Total Survivors |
| 37.7 | Overall Survival Rate (%) |
| 69.6 | Female Survival Rate (%) |
| 21.2 | Male Survival Rate (%) |
| 0.8051 | Model AUC |
| 75.5 | Model Accuracy (%) |
The dataset contains 600 passengers, of whom 226 (37.7%) survived. The logistic regression model achieved an AUC of 0.805 and an accuracy of 75.5%, indicating it can distinguish survivors from non-survivors substantially better than chance. Female passengers survived at 69.6% versus 21.2% for males, the most striking headline difference in the data.